2.2 



SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 65 



readings of the pyrheliometer simultaneously obtained. The deter- 

 mination of these secondary correcting factors and of the mean 

 bolometer constant for September 20 follows : 



Table 9 — Sensitiveness of Bolographic Apparatus 



.SB 



2 J3 



T 



II 



III 



IV 



V 



VI 



VII 



VIII 



X 



XI 



h m 



53.5 

 42.7 

 29.8 



16.9 



2.0 

 47-3 

 31.3 



10. o 



51.0 

 03.0 



17.15 



11.39 



7.71 



5. 75 

 4.46 

 3.67 

 3-o6 



2.53 

 1.62 

 1.38 



-2135 



2094 



2000 

 I96l 

 1887 

 1819 

 1807 

 1692 

 1624 

 1609 



5449 

 7663 

 9261 

 10613 

 1 1482 

 12276 

 12996 

 14005 

 15265 

 1 503 1 



.031' 



.000 



.003 

 .005 

 .003 



.000 



.994 

 .992 

 .005 



1 The correcting factor for holograph I is much above the usual magnitude. It was 

 not used for the following reasons: Firstly, the pyrheliometer exposes ^ hemisphere, 

 which is a sky area much larger than the sun. At ordinary air-masses the light of this 

 area of sky is negligible compared with sunlight. But at sunrise almost % of the solar 

 beam is lost by scattering in the sky, hence the light of the sky close to the sun is a very 

 perceptible fraction, perhaps 5 per cent, of that of the sun itself. Secondly the radiation 

 of the pyrheliometer to cold air and to space, which at high sun may reach nearly 0.005 

 calorie, is at the horizon counterbalanced by the radiation of the immense thickness of the 

 lower and warmer parts of the atmosphere, so that in comparison with high sun observations 

 the pyrheliometer reading at sunrise is probably about 1 per cent too high for this second 

 cause. Exact determinations of these corrections to pyrheliometry are proposed, but not 

 yet executed. Accordingly holograph I was omitted in the mean of column 10. 



2 Correction could not be determined because leaves of a tree intercepted the solar beam 

 during a part of holograph III. 



s Bolograph IX omitted, because shadow of a guy wire fell on the slit during a 

 considerable part of the time. 



In figures 3 and 4 we give plots to represent the results of the 

 spectro-bolometric observations of September 20 at different wave 

 lengths. The plots given in figures 3 and 4 are logarithmic. The 

 ordinates correspond to logarithms of the corrected heights of the 

 holographs at the 38 selected points, and the abscissae of the diagrams 

 represent the corresponding air-masses according to the tables of 

 Bemporad, corrected as heretofore explained. 



The original plots have been made on two different scales. In the 

 first, only those observations which we would ordinarily have used 

 for determining the solar constant of radiation were included. They 

 were plotted on the scales of ordinates and abscissae which we cus- 

 tomarily employ, in which, in general, 1 cm. = 0.01 in logarithm, and 

 1 cm. = 0.1 air-mass. In the other plot we have included all the 



